IndoorPlants

Background

This library is named for the wonderful coffee shop Tommy in Montreal where, as a good friend said: "the plants really tie the room together." It is my hope that this library will tie together pandas, sklearn, and matplotlib in ways that are useful for the practicing data scientist.

indoorplants provides data analysis and model validation tools, and the top-level namespace is broken out as such. Most of the code assumes that data is stored in DataFrame instances, and models are assumed to provide the sklearn interface. This code is built for Python 3.6+.

High-Level Description of Selected Functionality

On the analysis side of things:

• analysis.exploratory aids with exploratory data analysis
• analysis.wrangle is for data wrangling, including the detection of leaky features in classification problems
• analysis.features at the moment allows for calculating feature distances

On the validation side of things:

• validation.crossvalidate provides the core cross-validation functions
• validation.curves offers curves for ML model evaluation, including: validation, learning, and calibration

Example usage can be found in the census_data_analysis notebook.

Installation

You can run pip install indoorplants to get this repo.